1

Can civil society add value to budget decision-making?A description of civil society budget work

Warren KrafchikInternational Budget Project

1. The growth in independent applied budget work

Since the 1990s, civil society organizations have begun to play a larger role inthe budget process in some developing countries. This growth has beenparticularly rapid in the past few years and corresponds with the political trendtowards democratization. The first international meeting of civil society budgetgroups, convened by the International Budget Project in 1997, was attended by50 people representing budget organizations in fourteen developing countries. Bythe third conference, there were 100 participants from organizations in 25countries1. Most of the core groups from the first conference had deepened theirwork, while promising new efforts were represented from Ghana, Kenya, Russia,Croatia, Uganda, Peru and Indonesia.

The growth of independent budget groups is recent. While a few groups haveover 6 years experience, the majority of organizations have been engaged inbudget work for less than three years. Budgets remain closed processes indeveloping countries and most do not have independent budget analysis capacityoutside of the public sector. Nevertheless, applied budget groups are taking rootin a broad set of countries, including several not hospitable to inclusivebudgeting. Budget groups have been able to start despite extreme paucity ofdata, repressive regimes (Malawi), extensive corruption (Nigeria) and politicalturmoil (Indonesia).

The purpose of this paper is to examine this expanding contribution of civilsociety organizations to public budgets in developing countries. The material thatfollows will provide examples of civil society budget activities in a variety ofcountry contexts in order to measure the value of this work to public budgeting.

The growth of independent applied budget work organizations refers to a specificsubset of civil society organizations; hereafter called budget groups. The vastmajority of these groups operate independently of their country government andpolitical parties. It is this independence that often underlies the unique oversight 1 The work of the International Budget Project (IBP) is one of the few sources of information ofthis little documented trend. The Project was established by the Center on Budget and PolicyPriorities in Washington, D.C. in 1996 in order to help nurture independent budget organizationsin developing countries.

2

and information contribution of civil society to public budgeting. Eachorganization focuses on public sector budgets or government expenditure andtaxation issues. The work undertaken is applied or policy and impact oriented,highlighting the impact of public budgets on the poor. There is littleconcentration on developing theoretical models, although existing theoreticalwork is sometimes explained, adapted and or tested in application. Budget workis a way of conducting research on the budget that is of direct and timelyrelevance to current policy decisions. Groups strive to achieve analysis that istimely, accessible and accurate.

Despite great variation between budget groups, many share a common agendathat acknowledges the value of inclusive budget processes, independentoversight and a commitment to pro-poor choices within existing fiscalcapabilities.

Over and above these common points of departure, it is difficult to typecast thework of independent applied budget groups. Each group focuses on a widevariety of topics and employs several methodologies. Budget groups work ateither or a combination of national, state and local / municipal levels. Somegroups focus on simplifying the budget for popular consumption, some developexpertise in training, while others develop analytical or advocacy expertise orsome combination of these. Most groups monitor every stage of the budgetprocess, but often concentrate their interventions on a specific stage. Somegroups monitor the impact of the budget on the poor or develop methodologiesto monitor the welfare of specific interest groups such as women or children.

The origin of budget groups is similarly diverse reflecting the varied countrycircumstances within which budget groups operate. It also shows that there arefew limitations to the kind of organizations that can benefit from developingbudget skills. Some groups were established specifically to advance budgetissues, while the work of many has emerged to strengthen the work of anexisting organization. The majority of groups have emerged and operate withinthe non-profit, NGO / CBO sectors. However, several groups have beensuccessfully established within an academic environment and a couple of groupshave been initially supported from within the public sector. The leaders of thesegroups include activists, academics, former government leaders, business peopleand consultants.

Although not the central focus of this module, it is important to mention thatincreased civil society budget activities are often complemented by greaterindependent budget oversight within legislatures and by the increased criticalcoverage of budget issues in the media. A partnership between civil society andlegislatures is often the starting point for the development of local independent

3

budget work. In many cases, civil society has been able to build or enhanceresearch expertise that is lacking in legislatures, while legislatures can offeraccess to key moments in the budget process that are inaccessible to civilsociety.

2. International context

The role of legislatures in budgets is declining in Organization of Co-operationand Development member countries2. In contrast, the role of civil society andlegislatures in the budget is increasing in several developing and transitionalcountries. It seems only natural that civil society work has begun to gravitatetoward government budgets. The budget is the governments most importanteconomic policy instrument and should therefore reflect the nations priorities. Allpublic policies eventually have to confront the need for financial resources andcivil society groups can be more effective if they know about how the budget isdrafted, approved, implemented and evaluated. Yet, the timing of this growth inbudget groups reflects several international developments. These include:

Democratization: Budget work frequently flourishes in countries undergoing ademocratic transition such as South Africa, Russia, Uganda and Indonesia.Democratization often brings an associate increase in transparency andparticipation. This allows civil society and legislatures to play a moresubstantial role in public policy and simultaneously advocate for even greatertransparency. A virtuous cycle is thereby initiated, deepening democracy.

Decentralization: Democratization has often been accompanied by acomplementary process of decentralization that brings budgeting closer tocommunities. While decentralization may make it more complicated tomonitor budgets nationally, it may also open opportunities for greater citizenand local legislature involvement. In addition to possible formal opportunitiesfor involvement, given the weaknesses of many state and local institutions,decentralization often opens training and analysis opportunities for budgetgroups.

Public Expenditure Management: The progress of independent budgetinstitutions is taking place and, to some extent, incorporating revolutionarychanges to international public finance practices. New public expendituremanagement supports greater independent oversight capacity in civil societyand legislatures, although the terms of this engagement are not yet clear.

2 See Schick, Alan. (2001). Can National Legislatures Regain an Effective Voice in BudgetaryPolicy? Paper presented to the conference on Holding the Executive Accountable: The ChangingRole of Parliament in the Budget process, Paris, 24 25 January.www.oecd.org/puma/budget/journal.htm (forthcoming).

4

The new tools of outcome based medium-term budgeting are dependent ongreater transparency and more useful, timely and accessible data provision each of these important conditions for non-government budget work.Similarly, tighter standards for accountability enable enhanced externalmonitoring of public resources. It is interesting to note that the growth ofindependent budget organizations over the past two years has been mostrapid in Africa, particularly in the countries implementing new publicexpenditure management systems.

IFIs and poverty alleviation: Independent, applied budget work is alsoincreasingly supported by an emerging international consensus on thecomplementary role of the state and non-government actors in development.While the 1980s heralded the decline of communism and the victory ofprivate sector led development, international and regional developmentinstitutions generally acknowledge that progress against poverty andinequality has been less dramatic. Today, many of these institutions promotethe governance opportunities of public private partnerships. Civil societybudget work may represent one of the opportunities where the goals of thethe IFIs and civil society may be complimentary.

3. The case for civil society participation in the budget process

The growth of budget groups amounts may signal an important shift in publicfinance practice. In most countries, public budgeting has long been consideredthe exclusive preserve of the executive. It is only recently that the value ofopening budget processes to non-government input has been considereddesirable in some countries. For the majority of developing countries a set ofpowerful negative myths continue to constrain the independent budget work ofcivil society, legislatures and the media. These include:

Budgets must be formulated in secret or they may upset financial markets. Non-government intervention can destroy the integrity of the budget

envelope. Legislators and civil society have a greater interest in advancing the interests

of their constituents as opposed to the interests of the country as a whole. It is the governments mandate to produce the budget internally in a closed

process; and its prerogative for it to be rubber-stamped by the legislature.

A closer look suggests that many of these ideas are closer to myths. Budgetsecrecy may encourage market speculation, while greater transparency mayactually smooth market adjustment to known policy choices. Further, useful,accessible and timely budget information facilitates foreign and domestic privatesector planning and investment.

5

It is important to acknowledge the need for budget integrity and coherence isimportant as a guarantee of the prudence of the fiscal authority. This argumentmakes sense at the level of certainty about the fiscal envelope. But, budgetintegrity is unlikely to rely on the need for early certainty on the distribution ofresources between and within departments.

The tragedy of the commons is a danger of legislatures representing specificinterest groups too strongly, but it is not inevitable. A recent paper by AlanSchick, based on the experience of OECD countries, shows that greaterlegislature budget involvement can be compatible with outcomes that maintainfiscal discipline.

This conclusion could well hold for the work of applied budget groups. Whilesome groups may focus initially on expanding the budget envelope, as civilsociety budget engagement deepens, budget groups are more likely to focusreprioritization and the effectiveness and efficiency of expenditures. It does nottake groups long to realize that even if the size of the envelope is expanded,there are many more resources at stake in improving spending within theenvelope.

Finally, very few people would argue with the assertion that the government hasa mandate to produce the budget; but that does not mean it should entirelydominate a closed process. Moreover, there may be positive benefits forgovernment in accepting budget groups as a complementary player in fiscalpolicy. There will be disagreements, but over time budget groups may be able todeliver greater budget understanding and commitment from a broader group ofcitizens and better policy options.

Moreover, the relationship between civil society and government on budgetaryissues is not necessarily confrontational and the work of applied budget groups iscompatible with increased pubic sector budgeting capacity. Although it willdepend on the country context, many budget groups operate effectively ascritical allies of the government. Budget groups may strengthen governmentcapacity, for example, by providing training, undertaking research of interest togovernment, working with government in forums and building the potential ofthe legislature. Independence does imply the possibility of criticism wherenecessary and this can lead to confrontation, especially where other strategieshave not proved useful. But, by adjusting the tone and manner of criticalengagement and by producing accurate work, most groups are able to maintaina working relationship with the executive.

If government were to accept budget groups as a partner in national budgeting,there are ways to manage the potential negative implications. One practical

6

option is to separate discussions on the size of the budget from discussions ofprioritization. This is the case as is the case in Brazil, Switzerland and Sweden,amongst others. Government can also use engagement to focus and direct civilsociety to appropriate stages in the budget process. In a world where state andlegislature capacity is often less than desirable, these and other options offer anew form of public private partnerships that may enhance domestic budgetmanagement and outcomes.

In contrast to the above, there are several strong arguments for enabling civilsociety and legislature intervention in the budget. The work of budget groups,either directly or indirectly through the activities below, ultimately has thepotential to improve budget decision-making.

Simplifying the budget and deepening debate: Most budget groups begin orestablish as a core of their work developing budget literacy and debateamong citizens, interest groups and legislators. This augments the outreachcapacity of the state and legislature and builds understanding andparticipation among citizens. Simplified guides to the budget may cover thebudget process, the structure of the budget, budget trends and currentallocations, and some discussion of current issues.

Collating, synthesizing and disseminating budget information: These activitiessupport legislature and civil society inputs into the budget process. Theinformation may include the dissemination of cross-state or provincialinformation on social expenditures; or the identification and dissemination oflocal and international best practice.

Independent critical analysis: Civil society budget analysis can augment theresearch capacity of legislatures and the media and, sometimes, in theexecutive. It may provide one of the few sources of specialized data andanalysis on the impact of the budget on the poor. Budget group analysisincludes four major foci budget process or systems, general macro-economic, sector and group incidence and revenue analysis.

Bringing new information to budget decision-making: Civil society budgetgroups are often able to be in close, regular contact with citizens and interestgroups. This allows them to collate unique information on citizen priorities,non-government perspectives on the budget, expenditure tracking andproject impact analysis.

Training: Many civil society organizations develop budget training expertisethat is directed at augmenting the analytical and advocacy capacity of othercivil society organizations, legislatures and the media resulting in strongerinterventions and oversight. Several groups also provide training for

7

government departments and international institutions, such as the UnitedNations and regional trade organizations amongst others.

Building accountability: Through its analysis and dissemination activities, civilsociety organizations can assist citizens and the legislature to reinforcechannels of accountability. This may occur, for example, when groups provideaccessible information or pursue the findings of the auditor general withrelevant interest groups. It may also follow from monitoring or impactmeasurement activities that test or augment the data emerging from theauditor generals report.

4. Civil society in the budget process

This section is intended to further develop the discussion on the potential valueadded of civil society to the budget. It is written with the caveat in mind that thetrend to non-governmental applied budget work is recent. We are able to pointto the wide scope of applied work and some impacts, but this experience cannotyet be generalized. The evidence does point strongly however to the potential ofapplied budget work across a broad set of countries and despite significantobstacles.

We use the budget process as a framework of reference to organize ourdiscussion. This framework illustrates clearly that responding to the presentationof the annual budget in the legislature is usually the major focus of most budgetgroups activities during the year. Notwithstanding this fact, it is useful to bearthe following qualifications in mind.

The budget process itself is not divided cleanly into four stages. In practice,the stages overlap. Similarly, the work of budget groups usually spans acrossmore than one stage. Indeed, one of the constant success factors associatedwith groups around the world is that budget work requires full timecommitment. Budget groups must concentrate on all stages in the budgetprocess to be successful.

the budget process in each country context offers a unique set ofopportunities and constraints. Groups will choose how divide their energiesdepending on the relative opportunities at each point in the process.

Drafting stage

The drafting stage is the most closed part of the budget process in virtually allcountries and the most dominated by the executive. There is therefore usually

8

little formal opportunity for civil society input prior to the tabling of the budget inthe legislature.

The intervention opportunities for civil society that do exist in the drafting stageare often informal, based on the organizations initiative and networks, andreliant on departmental and ministerial contacts. For example, if it is known thata department or individual is concerned with a particular issue, this may createan opening to discuss relevant research. Alternatively, as the timing of thebudget process is generally known, a well-timed release of a research report maybe able to influence discussions in the drafting stage. Of course, civil societyorganizations may also rely on the media or other methods of capturing thegovernments attention during the drafting stage if it is considered that this willbe effective. The lack of formal opportunities for input does however mean thatit is more difficult to identify the impact of these activities on the executivesproposal.

One of the ways in which budget groups have taken the initiative to createopportunities in the drafting stage is to influence the set of priorities underlyingbudget policy choices. In Kenya, the Institute for Economic Affairs coordinates anannual civil society meeting with broad stakeholder attendance that culminates ina report providing government with a set of agreed priorities. The government inKenya has now formally requested this analysis as annual input into the draftingstage.

In a more elaborate process, the South Africa National NGO Coalitioncoordinated a two-year process of poverty hearings in many towns and citiesthroughout the country. The evidence of the poor was collated into a set ofpriorities by commissioners and delivered to government for inclusion in thebudget (Contact the Coalition through www.sanggoco.org.za).

Many organizations have tried to take pre-budget input a step further byproducing an alternative budget. Most of these attempts have failed due to thefinancial cost, data and modeling required to produce a comprehensive parallelbudget. Only two of these efforts, focusing on a defined theme or limitedsectors, have endured. These are the Alternative Federal Budget in Canada,prepared by the Canadian Center for Policy Alternatives(www.policyalternatives.ca) and alternative tax proposals produced by theInstitute for Fiscal Studies in United Kingdom.

There are two further government-led experiences of civil society engagement inthe drafting stage that may help to open formal opportunities for civil society inthe drafting process. The most famous example is the participatory budgetinginitiated and managed by the local government in the Brazilian town of PortoAlegre. The process was introduced when the Workers Party won the mayoral

9

race in 1989 on a participation ticket and has been replicated in approximately100 Brazilian towns with varying success. There is no precise model across thesetowns, but the programs are designed to incorporate citizens in the allocation ofresources, selection of broad social policies and monitoring of public spending.(For a review of participative budgeting in Brazil, seewww.internationalbudget.org/cdrom/papers/systems/ParticipatoryBudgets/Wampler.pdf)

The global trend towards fiscal devolution may provide further opportunities forlocal level budget participation. There are towns in South Africa, Peru, Russiaand Argentina amongst others that are currently experimenting with participatorybudgeting.

One of the success factors in participative budgeting worth mentioning in thecontext of this paper is that it may depend on a weak legislature or underminethe power of the legislature. It has proved successful in towns where the mayorhas virtually complete budgetary and administrative power and therefore thespace to innovate without the consent of the local legislature.

The scalability of participatory budgeting is not clear and currently the topic ofresearch in progress. Participatory budgeting has not yet proved workable at thestate or federal level in Brazil. There is also no example yet of a participatorynational budgeting process, although the Bangladeshi government hasintroduced an institutionalized, state led round of consultations conducted onlyone week prior to the tabling of the national budget (See the BangladeshInstitute for Development Studies www.bids-bd.org).

A recent Ugandan experience may point to a second potential opportunity forformal civil society engagement in the budget drafting stage in HIPC countries(Highly Indebted Poor Countries).3 When the National Resistance Movementcame to power in Uganda in 1986 it introduced a system of participatorydemocracy. This included participation of technically able stakeholders in a jointnational task force charged with prioritizing poverty reducing public goods insuccessive national poverty plans. The task force included civil societyorganizations, academicians, parliament, government and donors and based itsfindings on the collection of primary data and consultations with experts andservice providers.

3 The HIPIC initiative grants debt reduction relief to highly indebted poorcountries on the condition that funds released are deployed in poverty programsand that civil society is integrated into the drafting and monitoring of theseefforts.

10

With the formation of the HIPC initiative in the 1990s, the poverty eradicationaction plan was rolled over into the first PRSP (Poverty Reduction StrategyPapers). To comply with HIPC requirements, the Uganda government added abottom-up participatory poverty assessment project in 10 pilot districts to reviewthe priorities based on community level experience and develop mechanisms fortheir consistent involvement in poverty reduction strategies. Monitoring of thepoverty plan is undertaken biannually by the Finance Ministry and by civil societyorganizations monitoring the Poverty Action Fund established to disburse theresources released under HIPC (detailed under Implementation, below).

The early promise of the Ugandan model may be related to the unique politicalsystem. Participatory national priority setting has proved less successful inKenya, for example. However, the expanding PRSP process may present futureopportunities for early civil society involvement in setting budget priorities insome countries. The formal participation of the government, as the driver orpartner of the initiative is of course critical for success, as in the case ofparticipatory budgeting.

There is one further innovation that increases the scope for legislature and civilsociety input. Several countries have started publishing early details of macro-economic policy and a broad outline of the forthcoming budget in a pre-budgetstatement months before the tabling of the final budget. This has the effect ofseparating out discussions on macro-policy and the size of the budget fromdetailed interaction with department budgets in the legislature budget process.In some countries, such as Brazil, Sweden and Switzerland, the pre-budgetstatement must be debated and approved by the legislature (In Brazil, the rule isthis discussion and enactment must occur before representatives are allowed tostart the summer recess!). In other countries, such as Peru and South Africa,there is space for the legislature to debate but not vote on this broad statement.The benefit of the pre-budget statement is that it can protect the macro-envelope early in the process, while encouraging detailed engagement withissues of reprioritization and effective and efficient spending.

In sum, the drafting stage traditionally presents limited formal opportunities forcivil society to add value to the budget. At a local government level, participatorybudgeting may prove effective in environments outside of Brazil, but is unlikelyto do so at state or national level. At a national level, opportunity for influencethrough the setting of national priorities may grow through the expanding PRSPprocess. Nevertheless, civil society organizations do create their ownopportunities to influence the drafting process through the collation of usefulinformation on priorities and through the carefully direct or indirect contact withthe executive.

11

To the extent that budget groups are involved in the drafting stage, the benefitsof civil society involvement may be broader understanding of the choices facingbudget decision-makers and greater accuracy in identifying citizen prioritiesbetween elections.

The degree of civil society efforts in the drafting stage will be related to theintervention opportunities throughout the budget process, and particularly thelegislative stage. Involvement in the drafting stage may be particularly importantin countries where the legislative stage is undermined by weak or part-timelegislatures. In countries where there are few legislature opportunities foreffective and direct civil society intervention, groups may well strategize to directgreater resources to the drafting stage, despite its limitations.

Legislative stage

Many budget groups focus considerable energy on this stage for several reasons.In most democracies, the budget is tabled and must be approved in thelegislature. This is usually the first comprehensive sighting that civil society hasof the plan and a high profile opportunity to increase the impact of their work.The legislature process also offers the first formal opportunity for legislaturerepresentatives to debate and, in some cases, change the budget. This increasesthe opportunities for direct influence and the demand for NGO services such astraining and research. Moreover, in many countries, the legislature offers civilsociety direct intervention opportunities, such as public hearings.

Given parliaments role as the representative of citizens, this stage is oftenconsidered the most appropriate point for civil society intervention. However, theeffectiveness of civil society involvement in this stage depends on the strength ofparliaments own power in the budget process.

In theory, legislatures have significant power over the budget. The constitutionusually prevents funds being spent without legislative authorization. Legislaturedebate can be complemented with powers to summon expert witnesses andmembers of the government to give evidence. These powers are augmentedsignificantly where the legislature has the capacity to amend the budget. Inpractice, developing country legislatures are often undermined by severalfactors, including poor research capacity, the absence of powers of amendment,a weak committee system and the timing of the budget process.4

4 For a comparative analysis of legislatures and the budget, see Krafchik, W andWehner, J. (1998). Parliament and the budget. Available atwww.Idasa.org.za/final/publications/parliament1.htm

12

Despite these problems, several legislatures have recently taken the initiative toflex their budget powers. This is currently the case in Croatia, Kenya, Malawi,Mexico, Peru, South Africa, Uganda and Zambia, amongst others. Theweaknesses and strengths of legislatures and civil society budget groups cancombine to establish a potentially powerful oversight partnership. As theopportunities for greater legislature engagement with the budget expand, so toowill the opportunities for deeper civil society engagement with the budget. Inturn, civil society is able to augment the limited outreach and analysis capacity oflegislatures thereby adding value to the budget.

We turn now to the different ways in which budget groups add value to thelegislative budget process. The first major activity of budget groups at budgetpresentation time is the preparation of accessible summaries and guides to thebudget. In countries with limited budget analysis capacity in civil society or thelegislature, these products can serve as the first opportunity for independentoversight and participation. The recent work of the Institute for Public Finance inCroatia illustrates this point well.

The Institute recently published the first guide to the Croatian budget. Itprovides a general introduction to the budget and analyzes the majorexpenditures and revenue trends. It also covers extra-budgetary funds and theconsolidated budget of the general government and the different stages of thebudget process. The guide has been widely quoted, discussed and commentedupon in the Croatian media as it was published at the same time as theexecutives budget proposal for the next financial year was being discussed inparliament. In a comment underscoring the usefulness of the guide, one of themembers of the parliament stood up, flashed the guide, and said to the deputyminister of finance: "Now we don't have to (only) listen to you anymore, wehave a (alternative) guide!"

Budget training is a second major activity of budget groups during the legislaturestage. These programs may be targeted at legislatures, media and civil societyorganizations, as well as government personnel. CIDE in Mexico is one exampleof a group providing comprehensive budget training to these target groups. Thethree-day budget course on the federal budget has the objective of teaching thebasic knowledge needed to carry out an analytical view of Mexicos budget. Thecourse covers the broad contours of Mexicos federal budget, its origin, approvalprocess, administration, and impact. This training became an important resourcefor legislators when the PRI lost its majority in Congress and the potentialexisted for a dramatic increase in the role of the legislature in the budget (Seewww.cide.edu).

Training may also be coupled with more intensive technical support. Idasa inSouth Africa formed a partnership with a multi-national accounting firm to

13

provide training and technical support to provincial legislatures in South Africaduring the legislature budget process. The project creates a pool of trainedanalysts from both organizations to provide general training and in-depth supportto committee chairs for the month-long legislature debating period. For furtherexamples of budget training see the work of DISHA in India training dalits(contact Disha@vsnl.net.in) and the work of INESC in Brazil (inesc.org.br).

The third major activity of budget groups during the legislative stage involves thepreparation and coordination of independent analysis. Given the limited fiscalanalysis capacity outside of the private sector in developing countries, thisconstitutes one of the few accessible, timely, critical interpretations of thebudget. It is often the only available analysis on the implications of the budgetfor low-income communities.

Analysis of the executives budget proposal takes four major forms:

A few groups produce a comprehensive analysis of the budget, such as theresponse of the Consumer Unity and Trust Society in India to the 2000/01Indian Union budget. The analysis argues for government to stick with itscommitment to phase out non-merit based subsidies in favor of merit basedsubsidies especially in education and health. The memorandum suggests thatthe resources released through this shift should be reprioritized towardsprograms addressing child labor issues (Seewww.internationalbudget.org/cdrom/papers/analysis/CUTSINdiabudget%202000.htm).

A complementary approach is provided by the Catholic Commission for Justiceand Peace in Zambia. The Commission monitors social service delivery duringthe year through a set of monitors around the country and combines thisinformation, together with the budget comments of a focus group of NGOs,to draft a broad comment on the poverty impact of the budget. The post-budget statement covers health, welfare, education, agriculture, debt, taxand monetary policy. (See www.ccjp.org.zm/2001pre_budget.htm)

Another complementary approach is a broad survey of budget transparencyand participation in the budget process that was designed by theInternational Budget Project and Idasa and tested in South Africa. This iscurrently being applied in four countries (Ghana, Kenya, Nigeria and Zambia)and five Latin American countries (Argentina, Brazil, Chile, Mexico and Peru).A detailed description of the survey and the South African results is presentedin Case study 2 that is appended to this document.

A few groups analyze a particular macro-economic issue. For example, groupsmay focus on the macro-economic parameters underlying the budget or the

14

effect of the budget on macro-variables such as the deficit, national debt,interest rates, growth or employment. In its submission to parliament on the2000/01 budget in South Africa, Idasa questions the governments optimisticgrowth forecast and traces the impact of the likely slower growth onprograms for the poor (Seewww.Idasa.org.za/final/publications/bisbudget2000.sub.htm)

A large proportion of groups focus analyze the impact of the budget onparticular sectors or socio-economic groups, such as low-income people, theelderly or racial or ethnic groups. Often the sector foci are on health, welfareor education given that these comprise the majority of social services.Analyses consider sector allocations in relation to allocations to other sectorsor to the entire economy, or in relation to historical levels of support.

As government budgets do not generally reference the effects on particularsocio-economic groups, as opposed to departments, several budget groupshave developed creative methods to undertake incidence group analysis.Work on gender budgets has become a popular area of activity in many partsof the world as shown in a recent paper by Debbie Budlender5. The TanzaniaGender Networking Program established their gender budget initiative in1997. It focuses on macro-economic policies and planning with a stress onbudgets and budgetary processes and their impact on men and women inTanzania. (See Tastes of Success, p 53).

Work on childrens budgets has also advanced recently in several countries.The South African Childrens Budget was launched in 1998. The projectexamines what resources the government is allocating to childrens programsand whether these programs adequately reflect the needs of children asestablished by the United Nations Convention of the Rights of the Child. (TheSouth African Childrens Budget is accessible atwww.idasa.org.za/bis/children)

As a fourth foci, budget organizations are increasingly undertaking taxationand revenue analysis. This work may challenge the distributional impact ofthe proposed taxes or the suitability and sufficiency of revenue proposals tofund proposed expenditures. Tax administration and compliance are furtherissues pursued. The Center on Budget and Policy Priorities recently played anactive role in deepening the recent tax debate in the US. In a paper onPresident Bushs tax cut proposal, the organization argues that the cut maynot help the U.S. out of its current economic slowdown (Seewww.cbpp.org/2-22-01tax.htm).

5 See Debbie Budlender .(2001). A Review of Gender Budgeting. Commonwealth Secretariat.

15

The Institute for Public finance in Croatia is similarly working on responses tothe governments current reform of the overall fiscal system, including taxadministration reform. In a recent paper, they question the extent to whichthe proposals for tax reform are in line with the theory and practice of taxadministration in developed countries and countries in transition. (Seewww.ijf.hr/ocpapers/op-5.htm)

In addition to their own analysis, some budget organizations facilitate or hostforums to collate budget analyses of other civil society and community groups.This assists in providing government with a rich source of information from abroad group of popular interests. A number of civil society organizations inZambia facilitate commentaries following the presentation of the budget in thenational parliament. One of these forums is the annual National BudgetWorkshop organized by the Economic Association of Zambia that brings togethercivil society groups with the Minister of Finance and Economic Development.Contributions have covered issues such as macro-economic policy, revenuemeasures, expenditure patterns and the gender impact of the budget.

In sum, the legislative stage of the budget process often offers budget groupsthe strongest opportunity for formal involvement and leverage. In these cases,civil society work can add maximum value by offering the full spectrum ofactivities including training, budget guides and analysis. Budget work focussingon the impact of the budget on the poor can provide a unique resource, augmentthe paucity of trained analysts in the public sector and potentially improve policychoices.

The contributions of budget organizations to the legislature stage does howeverdepend significantly on the degree of opportunity for legislature budgetinvolvement. Where this is limited budget organizations can still add value to thebudget by simplifying and clarifying the budget proposals, providing training andby working closely with legislators. However, in these situations budget groupsare likely to channel a greater proportion of their resources to other stages in thebudget process, including the drafting process.

Implementation stage

Implementation is primarily an executive function. For most applied budgetgroups, the importance of this stage is the opportunity it holds for collatinginformation on expenditure, revenue and outputs that will contribute to thequality of their participation in the legislative stage. However, weak publicimplementation capacity and accountability mechanisms are driving budgetorganizations to focus on implementation issues.

16

Organizations in India, such as the Center for Budget and Policy Studies inBangalore (contact CBPS@vsnl.com), have produced useful analysis of thedifference between planned and actual expenditures. Data improvements inSouth Africa are also enabling this type of analysis for the first time. The biggestobstacle to this work in other countries is the paucity of government data,including timely and accurate information on disaggregated actual expenditure,government borrowing and debt redemption; as well the capacity to provideincidence benefit reports.

Ritva Reinikka at the World Bank has helped pioneer methods to measure thedifference between planned and actual expenditure where data is scarce. Thiswork relies on new community level data sources in Ghana, Uganda, Tanzaniaand Honduras and may be adapted by independent budget groups in the future6.

Uganda provides an interesting experiment in implementation monitoring thatmay open opportunities in other countries. In 1998, the Ugandan governmentestablished a Poverty Action Fund (PAF) as a mechanism to target, protect andmonitor funds released by the HIPC initiative and donors for poverty programs.The PAF does not operate as an extra-budgetary fund, but is integrated into thebudget. The Fund involves civil society organizations in the selection of projectsand in monitoring the impact of PAF expenditures by allocating 5% of the fundto sustain monitoring activities. Monitoring is coordinated by an applied budgetwork coalition, the Uganda Debt Network, and is undertaken through periodic(quarterly) field surveys by a team of researchers and community members. Theinitiative also tracks monthly expenditure releases from central government tolocal governments and reconciles these with releases from the central bank.Quarterly progress reports are presented at multi-stakeholder meetings. In arelatively short period of time, this initiative has helped to identify the blockagesin funding to local government, increase funding to poverty relief programs andshifted expenditures towards identified priority sectors. (See www.udn.org.ug/)

There are many civil society organizations with experience in program and citizen/ customer evaluation. The work of the Public Affairs Center in Bangalore isparticularly well known in this regard (www.pacindia.org). There is also civilsociety work in Ghana and Zimbabwe on tracing the impact of structuraladjustment programs through community surveys. But, there are few examplesof organizations that track budgetary flows through to program impact. Workalong these lines is only beginning. Idasa in South Africa is undertaking two pilot

6 See for example, Emmanuel Ablo and Ritva Reinikka - Do Budgets reallymatter? Evidence from Public Spending on Education and Health in Uganda.Available from Kathryn Rivera at the World Bank krivera@worldbank.org)

17

projects tracking the impact of pubic expenditures on special courts for childrenand rape victims. SAPRI in Ghana and the Catholic Commission for Justice andPeace in Zambia have used their structural adjustment monitoring work as aspringboard to focus more directly on budget flows and have plans to test thelink between budget plans and disbursements. The latter has an extensivecommunity network that has been successfully used in the past to trackblockages in the flow of public funds from provincial to local level.

In sum, it is too early to evaluate the replicability and sustainability of theUgandan poverty fund monitoring initiative - but, these efforts cannot substitutefor government information collation capacity. Civil society can help to someextent but this requires a rare partnership of sophisticated policy NGOs withstrong community partnerships. As data improves, applied budget groups will beable to provide more sophisticated checks on the congruency and impact ofexpenditure. This opportunity is unlikely to emerge as a significant space forbroad civil society engagement, but might work in generating data for a definedset of poverty expenditures or as a check on government generated numbers.

At present, the primary importance of the implementation phase for most budgetgroups is the opportunity to collect data on implementation to use as evidence inthe legislative and auditing stage. As argued by the Catholic Commission forJustice and Peace in Zambia: The futility of our system of commenting on thebudget once it has been presented is that no influence can be made on t hecontents of the budget. Not even the debate by MPs influences the budget underdiscussion. Our post-budget challenge therefore remains at monitoring theimplementations of commitments made in the budget throughout the year,drawing lessons that will be vital in our submissions to the next budget (CCJP,2000 Post Budget statement).

Auditing stage

The primary constitutional actors in the auditing stage are the Auditor-General orCourt of Accounts and the Legislature. These agencies are usually required toreport to the legislature as keeper of the purse. The physical and humanresources required for a full state or federal audit limits civil society interventionsto checking the Auditor-Generals findings in discreet areas. The majoropportunities for independent budget organizations follow from the findings ofthe audit and the legislature procedure to evaluate these.

This may be one area where applied budget organizations have not sufficientlyexploited the potential opportunities. Where developing country auditing iscompetent, these reports often document a litany of mis-expenditure, mal-expenditure and procurement irregularities. However, few of the organizations in

18

the International Budget Project network dedicate resources to monitoring andfollow-up on the audit function.

There are several reasons to account for the lack of civil society audit stageactivity. In most developing countries, the audit report is presentedapproximately two years after the close of the financial year. The audit report isswamped and often confused by current budget issues. In parliamentary systemsin particular, the absence of effective legislative budget powers provide littleincentive to scrutinize whether the funds have been spent as authorized. The lowinterest in following-up on the audit report reinforces this tendency. This isfurther compounded by the split between committees dealing with publicaccounts oversight and those dealing with budget analysis and authorization.

The experience with participatory auditing in Kerala, India points to potentialfurther oversight opportunities for a few civil society groups during the auditstage. The first is the development of participatory audit techniques atcommunity level in India. This fascinating practice involves the local auditormeeting with village members to physically audit the outputs planned andapproved in the local budget (See the work of Anne-Marie Goertz and RobJenkins at the Institute for Development Studies at the University of Sussex).

5. Summary and conclusions

Applied budget groups have recently begun to expand the role of civil society inthe budget. Their interventions have enabled broader understanding andparticipation in the budget process and more effective oversight. In part, thesebenefits occur indirectly through other organizations or legislatures as a result oftraining and dissemination activities that deepen public debate and expand publicparticipation. However, some evidence is emerging that their activities can alsodirectly improve budget decisions and outcomes.

Many groups are still developing skills and organizational capacity. Only a smallset of experiences already point to significant impact. In Mexico, the work ofFUNDAR in the 1990s halted the use of a Presidential secret account that did notrequire the approval of Congress. In Israel, the work of ADVA pressured thegovernment to retract changes to the welfare benefits that would have beenprejudicial to the countrys poorest families and elderly citizens. In South Africa,the work of Idasa led to an improvement in a new basic financial managementbill through the inclusion of clauses on Virement7 and sanctions formismanagement. In Russia, the work of Volgo-Vyatsky contributed to eliminating

7 Virement refers to the extent to which a government department can spend in a way that is notcongruent with budget plans without reverting back to the legislature for approval.

19

a corruption-prone arrangement that allowed firms to pay a portion of theirtaxation in goods and services.

The legislature stage usually holds the most opportunity for budget groups. Thisstage is geared to benefit from the full spectrum of civil society budget research,training and dissemination activities and also provides a forum for the formalinclusion of the sector into the policy process. At present, as far as the budgetdrafting and implementation phases are concerned, the participation of budgetorganizations is largely limited to collating information on priorities andimplementation. It is still too early to assess the full implications of the promisingparticipation efforts associated with the PRSP process or the local governmentparticipatory budgeting experience. Given resource implications and technicalrequirements of community surveying and impact evaluation, it is likely thatinvolvement in these stages will be limited to a few organizations evaluating asmall proportion of the budget or particular projects. Conversely, there isprobably greater capacity for civil society involvement than currently exploited inthe auditing stage. There is an important role in tracing, with legislatures, thefindings and recommendations of the auditor-generals report. As far as a broaderrole in the auditing stage, civil societys role is probably limited to cross-checkingparticular projects, or areas of concern rather than duplicating the role of theauditor-general.

We noted earlier that applied budget organizations may add value to budgets inseven ways. The primary value added in the drafting stage may be their ability tobring new information to the public debate on citizen priorities and, throughtraining, building the capacity of communities to take part in this process. Duringthe legislative stage, the contribution of applied budget organizations includesbuilding budget literacy and training and analysis that brings a pro-poorperspective to budget deliberations. This is the stage where the analytical skillsof applied budget groups are most evident and where their ability to improvebudgetary decision-making is maximized. During the implementation phase,budget groups may help to a limited extent in collating information on programimpact. In addition to similar evaluation activities in the audit stage, appliedbudget groups may be able to play a bigger role in helping legislatures tomonitor the impact of the official audit and in interpreting and disseminating thefindings of the auditor-general.

These early results of the rise of budget groups provide cause for optimism.Groups have been established in diverse settings and shown resilience tosignificant political and data obstacles. There seem to be few organizations thatcannot benefit from budget analysis skills. Consequently, we can expect thegrowth in budget work to continue to broaden and deepen in the developingworld. Part of this growth may well take the form of greater interaction between

20

budget groups and other civil society organizations, including human rightsorganizations and activities.